Context. Imaging surveys of dust emission at (sub)millimetre wavelengths provide a powerful tool for studying molecular clouds and the early stages of star formation. Aims. Through submm dust continuum mapping, we attempt to search for genuine infrared-dark clouds (IRDCs) and precursors to massive stars and stellar clusters in the Galactic plane, and to determine their basic physical properties. Methods. We have mapped four selected fields of about 0.°5×0.°5 that contain Spitzer 8-μm dark regions with LABOCA at 870 μm. Selected positions in the fields were observed in C17O(2−1) to obtain kinematic information. The obtained LABOCA maps are used in conjunction with the Spitzer IR images. Results. The total number of clumps identified in this survey is 91, out of which 40 (44%) appear dark at 8 and 24 μm. The remaining clumps are associated with mid-IR emission. Seven clumps associated with extended 4.5 μm emission are candidate extended green objects (EGOs). Filamentary dust “ridges” were found towards the Spitzer bubbles N10/11 in one of our fields. The relative number of IR-dark and IR-bright clumps suggests that the duration of the former stage is about 1.6 × 105 yr. The mass distribution of the total sample of clumps, and that separately constructed for the IR-dark and IR-bright clumps, could be fitted at the high-mass end with the power-law function dN/dlog M ∝ M−Γ, where Γ ≃ 0.7...0.8. The C17O observation positions appear to be dominated by non-thermal motions, and the data also revealed some potential sites of strong CO depletion. In G11.36+0.80, which is the best example of a filamentary IRDC in our sample, the clumps appear to be gravitationally bound. The fragmentation of the filament can be understood in terms of a sausage-type fluid instability, in agreement with the results for other IRDCs. The fragmentation and the CO depletion timescales in G11.36 appear to be very similar to each other. Conclusions. Many of the identified clumps are massive enough to allow high-mass star formation, and some of them already show clear signposts of that. In the N10/11 bubble environment, the morphology of the detected dust emission conforms to the triggered high-mass star formation in the system. The clump mass distributions are similar to those found for diffuse CO clumps, and can be explained by the action of supersonic turbulence. The formation of filamentary IRDCs might be caused by converging turbulent flows, and the same process may play a role in exciting the fluid perturbations responsible for the fragmentation of the clouds into clumps.